Update README.md

README.md

@@ -1,26 +1,33 @@
----
-license: mit
-dataset_info:
-  features:
-  - name: id
-    dtype: int64
-  - name: url
-    dtype: string
-  - name: title
-    dtype: string
-  - name: text
-    dtype: string
-  - name: words
-    dtype: int64
-  splits:
-  - name: train
-    num_bytes: 2354942374
-    num_examples: 399131
-  download_size: 1107142059
-  dataset_size: 2354942374
-configs:
-- config_name: default
-  data_files:
-  - split: train
-    path: data/train-*
----
+---
+license: mit
+dataset_info:
+  features:
+  - name: id
+    dtype: int64
+  - name: url
+    dtype: string
+  - name: title
+    dtype: string
+  - name: text
+    dtype: string
+  - name: words
+    dtype: int64
+  splits:
+  - name: train
+    num_bytes: 2354942374
+    num_examples: 399131
+  download_size: 1107142059
+  dataset_size: 2354942374
+configs:
+- config_name: default
+  data_files:
+  - split: train
+    path: data/train-*
+---
+
+# This Dataset 
+the akhooli/arwiki_128 dataset is a sample of the 2023 Arabic Wikipedia dataset originally published by Wikimedia (wikimedia/wikipedia).  
+The full Ar wiki is around 1.2 million articles while this dataset is around 400k with article length in the range 128-8000 words.  
+Words were counted using a simple method (space seperated).   
+Please use the [original](https://huggingface.co/datasets/wikimedia/wikipedia/viewer/20231101.ar) full dataset. 
+This sample is being used for continued pretraining of some multilingual LLMs for testing purposes and may be removed later (do not reference in published work).